Fringe evaluation and phase unwrapping of complicated fringe patterns by the data-dependent fringe processing method

Evaluation of noisy fringe patterns is important for solving problems of nondestructive testing and is widely used in moire, holographic, and speckle interferometry. In this paper, a new method of fringe pattern analysis has been proposed based on iterative estimation of local fringe amplitude, spacing, and orientation. Fringe parameter evaluation at each iteration step allows the formation of a two-dimensional (2-D) data-dependent anisotropic impulse response of a spatial filter that allows suppression of the noise influence without decreasing the fringe visibility. When the local fringe parameters are obtained, it is possible to find the phase difference at each point of the fringe pattern and at any other one within the local 2-D area. These local phase differences are utilized to recover the wrapped fringe phase at the point using a local model of the fringe pattern. fitted under the criterion of root-mean-square error minimization. The high noise immunity of the proposed method was verified experimentally when processing complicated fringe patterns with 2-D fringe phase unwrapping.